White being equalish is no small achievement in this line.

White can improve with 10.0-0 c4 11.dxc4 dxc4 12.Qx8+ Nxd8 (Kxd8 is unclear) 13.Nxb5 cxb3 14.Nc7+ Kd7 15.Nxa8 dxc2 16.Rf2 Bb7 17.Rd2+! Kc8 18.Rxc2+, probably with some advantage. The difference is that Black doesn't have 17...Nd5. It seems that this stems from Bulgarini.
Dolzjenkov's piece sac looks great. So yes, Black has to look at 10...d4 and 10...h6. As 10...d4 11.Ne2 Na5 12.Bg5 just transposes I suppose we have to look at 12.Qe1 Nxb3 13.axb3.
Still 10...h6 might be best as a winning attempt.

I am slightly puzzled by a line that is dealt with by Charlie Storey in his recent book "The Sniper" (Everyman 2011) and that is also mentioned by Gawain Jones in his book "Starting Out: Sicilian Grand Prix Attack" (Everyman 2008):

1.e4 c5 2.Nc3 Nc6 3.f4 g6 4.Nf3 Bg7
Obviously a position that could as well have arisen via a Sniper move order such as 1.e4 g6 etc.
5.Bc4 e6 6.f5 Nge7 7.fxe6 fxe6 8.d3 0-0 9.0-0 d5 10.Bb3 Na5
10...Na5 is awarded an exclamation mark by Storey, who comments: "An important moment as the knight prepares to neutralize any aggressive potential of the bishop."
10...Na5 is the most popular move in this position according to Megabase 2011, but apparently Black has two decent alternatives at his disposal, viz. 10...d4 and 10...h6; both of them score slightly better than 10...Na5. 
11.Bg5 d4
"Shutting the centre is nearly always good if you gain some initiative with it." (Storey)
12.Ne2
Neither Storey nor Jones comments on this move ...
12...h6 13.Bd2 Nec6
13...Nxb3 14.axb3 e5 15.b4! with counterplay for White according to Storey.
14.Qe1 Nxb3 15.axb3 e5.
More or less with equality, it seems. Ristoja (2317) – Inarkiev (2604), Izmir 2004, a game quoted in both books.

Can't argue with that, can you?

Well, Megabase 2011 has eleven games reaching the position after 11...d4. In eight of these games White played 12.Ne2, in two games he opted for 12.Nb1. Neither of theses moves looks really dangerous for Black.

However, in the game Dolzhenkov (2341) - Stepanov (2244), Rybinsk 2001, White came up with the stunning piece sacrifice 12.Nd5!? Oops ...
That game continued with 12…exd5
What else? 12…Nac6?! 13.Qe1!± with the idea 14.Qh4 [Rybka 4, Houdini 1.5].
13.Bxd5+ Kh8 14.Qe1
White wants to play Qh4. Note that 14.Qe1 comes with tempo; as the black queen is overloaded (Na5, Ne7), White is threatening Bxe7 now.
14...Nac6?!
Some alternatives:
a) 14...Rf6?! 15.Ne5 Qb6 16.Bxf6 Bxf6 17.Qg3± (intending either Rxf6 followed by Rf1, or Nf7+ followed by Nd6[+]);
b) 14...Bf6? 15.Bxf6 Rxf6 16.Qh4 Nxd5 17.Ng5! Qe7 18.exd5+- (19.Rae1 is coming);
c) 14...Re8? 15.Qh4 Qc7 16.Bf4 Qb6 17. Bf7+-;
d) 14...Qd6 (or 14…Qc7) looks relatively best, but after 15.Bxe7 Qxe7 16.Qxa5 Black is simply a pawn down for not much compensation, isn't he? Check for yourselves ...
15.Bxc6 bxc6 16.Qh4 Re8 17.Ne5! Qc7
17...Bxe5 18.Rf7 Bg7 19.Bf6+-
18.Bf6
18.Rf7!±
18...Bf5?
18...Nf5 19.exf5±
19.Nf7+?
19.exf5+-; 19.Rae1+-
19...Kg8 20.Nh6+ Bxh6 21.Qxh6 Nd5 22.exd5
22.Rxf5!±
22...cxd5?
22...Qf7 23.Qh4±
23.Rxf5!+- gxf5 24.Qg5+ Kf8 25.Rf1 Qf7 26.Qh6+ Kg8 27.Rxf5 1-0  

In the light of this nice attacking game, the move order 10...Na5 11.Bg5 d4 somehow does not look ideal. Better may be 10...d4 11.Ne2 (11.Nd5 doesn't work here, of course) and only now 12...Na5, when 13.Bg5 transposes to Ristoja – Inarkiev. White has an alternative in 13.Qe1 though, which, however, Black may also be able to deal with.

There's a slight mystery about the game Dolzhenkov – Stepanov. Has it really escaped Storey's and Jones' (or, for that matter, any other author's) attention? Or is 12.Nd5 not as strong as it appears, after all? Am I missing something?

Any thoughts?